RESOLVED FIXED 211291
Sending WebRTC network packets should not go through the main thread 
https://bugs.webkit.org/show_bug.cgi?id=211291
Summary Sending WebRTC network packets should not go through the main thread
youenn fablet
Reported 2020-05-01 07:33:07 PDT
Sending WebRTC network packets should not go through the main thread
Attachments
Patch (3.17 KB, patch)
2020-05-01 07:36 PDT, youenn fablet 		no flags
DetailsFormatted DiffDiff
Patch (16.06 KB, patch)
2020-05-05 08:17 PDT, youenn fablet 		no flags
DetailsFormatted DiffDiff
Patch (16.07 KB, patch)
2020-05-05 08:26 PDT, youenn fablet 		no flags
DetailsFormatted DiffDiff
Patch (25.77 KB, patch)
2020-05-06 05:49 PDT, youenn fablet 		no flags
DetailsFormatted DiffDiff
Patch (21.43 KB, patch)
2020-05-06 07:42 PDT, youenn fablet 		no flags
DetailsFormatted DiffDiff
Patch (25.76 KB, patch)
2020-05-06 07:48 PDT, youenn fablet 		no flags
DetailsFormatted DiffDiff
Patch for landing (26.20 KB, patch)
2020-05-06 10:22 PDT, youenn fablet 		no flags
DetailsFormatted DiffDiff
Patch for landing (25.76 KB, patch)
2020-05-07 01:56 PDT, youenn fablet 		no flags
DetailsFormatted DiffDiff
Show Obsolete (7) View All Add attachment proposed patch, testcase, etc.
youenn fablet
Comment 1 2020-05-01 07:36:00 PDT
Created attachment 398182 [details] Patch
Eric Carlson
Comment 2 2020-05-01 09:29:29 PDT
Comment on attachment 398182 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=398182&action=review > Source/WebKit/WebProcess/Network/webrtc/LibWebRTCSocket.cpp:61 > void LibWebRTCSocket::sendOnMainThread(Function<void(IPC::Connection&)>&& callback) Should this be renamed now that it doesn't send on the main thread?
youenn fablet
Comment 3 2020-05-01 09:41:25 PDT
(In reply to Eric Carlson from comment #2) > Comment on attachment 398182 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=398182&action=review > > > Source/WebKit/WebProcess/Network/webrtc/LibWebRTCSocket.cpp:61 > > void LibWebRTCSocket::sendOnMainThread(Function<void(IPC::Connection&)>&& callback) > > Should this be renamed now that it doesn't send on the main thread? Right, we also need a better handling of getting the IPC::Connection. The callOnMainThreadAndWait should be removed to ensure we do not end up in a deadlock.
youenn fablet
Comment 4 2020-05-05 08:17:31 PDT
Created attachment 398508 [details] Patch
youenn fablet
Comment 5 2020-05-05 08:26:31 PDT
Created attachment 398509 [details] Patch
youenn fablet
Comment 6 2020-05-06 05:49:13 PDT
Created attachment 398604 [details] Patch
youenn fablet
Comment 7 2020-05-06 07:42:21 PDT
Created attachment 398612 [details] Patch
youenn fablet
Comment 8 2020-05-06 07:48:09 PDT
Created attachment 398613 [details] Patch
Eric Carlson
Comment 9 2020-05-06 09:44:32 PDT
Comment on attachment 398613 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=398613&action=review > Source/WebCore/ChangeLog:12 > + Fix a case where the SDP would be badly formatted we do not have yet a MDNS name for the corresponding IP address. s/would be badly formatted we do not have yet a MDNS name/would be badly formatted if we do not yet have an MDNS name/ > Source/WebCore/ChangeLog:13 > + Small refactoring to use less else if s/use less else if/use early returns/
youenn fablet
Comment 10 2020-05-06 10:22:08 PDT
Created attachment 398634 [details] Patch for landing
youenn fablet
Comment 11 2020-05-07 01:56:28 PDT
Created attachment 398714 [details] Patch for landing
EWS
Comment 12 2020-05-07 02:23:48 PDT
Committed r261277: <https://trac.webkit.org/changeset/261277> All reviewed patches have been landed. Closing bug and clearing flags on attachment 398714 [details].
Radar WebKit Bug Importer
Comment 13 2020-05-07 02:24:16 PDT
<rdar://problem/62967836>
Note You need to log in before you can comment on or make changes to this bug.